Haxby et al. (2001) recently made an important contribution to our knowledge of the neural bases of visual categorization. While undergoing fMRI, participants viewed photographs of exemplars from 8 common object categories. The researchers observed distributed and overlapping category-related activity patterns in the ventral temporal cortex. These patterns were so robust that the category of stimuli participants were viewing could be predicted with almost perfect accuracy by examining the activity in the brain. In the present proposal, a series of fMRI studies will follow-up on two aspects of Haxby et al.'s findings. First, the topographic organization of ventral temporal category representations will be examined in light of a new theory (the Similarity-in-Topography principle; Simmons & Barsalou, 2003). According to the SIT principle, as the exemplars of two object categories become more similar visually, the distributed patterns of ventral temporal cortex activity associated with viewing those two object categories should become more similar topographically. If this is the case, then the topographic similarity of the ventral temporal category representations should be predicted by the visual similarity of the object categories. Experiment 1 will address this hypothesis. In addition, Experiments 2 and 3 will examine the stability of ventral temporal category representations with respect to changes in research method (blocked vs. random presentation of category exemplars) and taxonomic context (do category representations change with changes in the taxonomic relatedness of the surrounding categories?). [unreadable] [unreadable]